Comparison of soil erosion under no-till and conventional tillage systems in the high rainfall Mlondozi area, Mpumalanga province, South Africa

Abstract

Rural agriculture in Mlondozi, as for South Africa, is has a low productivity, which is the result of poor knowledge, information, beliefs and land tenure which limits the acquiring of loans for inputs. A LandCare Project was conducted for four years. Training was in the form of farmer managed research demonstrations which included the taught Conservation agriculture farming system which they compared to the Traditional farming system. Eighteen farmers initially joined the program and their soils were monitored for four years. At the end of the project undisturbed soil samples were taken from their fields where maize was cultivated following no-till (NT) farming system, and the conventional tillage system (CT). The soils from the two farming systems were compared using a laboratory rainfall simulator for run-off, erosion and infiltration. Each storm event in the rainfall simulator lasted for a period of 110 minutes (50 rotations). There were two statistical analyses done on the results. The first was a t-test was applied to the data to test for differences between the two systems, with a sample size of 72, at 18 sights with 4 replicates, except for carbon which was 36 analyses for the 18 sights. There was a significantly higher soil loss for NT soils for storm 1, compared to the CT soil, and a non significant difference for run-off for storms 1 and 2. Infiltration was significantly lower for the NT soil for the first storm, and not significantly higher for the NT soil for the second storm. The CT soils had a significantly higher infiltration rate for the first 16 rotations. After 68 minutes (rotation 34) the NT soils infiltration rate was higher. For simulated storm 2 the CT soils had a slightly higher infiltration rate up to 32 minutes (16 rotations) where after NT soils had a higher infiltration rate. Between 48 and 80 minutes (rotations 24 and 40) the NT soils had a significantly higher infiltration rate. From the results it can be concluded that the NT soils maintained the aggregate stability far longer than the CT soils. The initial and final infiltration rates were compared for the NT and CT soils for the simulated storms 1 and 2. The CT soils’ initial and final infiltration rate was similar, while the NT soils had a higher initial and final infiltration rate for the second storm, due to the soils settling with the first storm. The results question current literature that states that sealing of soils is a permanent feature. The carbon content of the NT soils was not significantly higher than the CT soils, which corresponded with the results.